 I think your last sentence sums it all. I think we are the most hated panel in the room now, between you and the lunch. But we'll try to make it worth for you. So thank you so much for joining us. Thank you so much all the panelists joining in today. I think we all talking about Generative AI since morning and maybe since weeks or since months. And it's surprising to know that I think it's a decadal moment for all of us. It just a little over one year since Generative AI has arrived. I think the transformation and the recall and the everything to do around it is being really impressive. I think cloud took almost 10 years to make that kind of waves and JNEI just took almost not even a year to make those kind of waves. So I think, so with that, I think in this panel, we will be talking about real use cases. How are we using Generative AI from hype to reality? I think that's one area we will touch upon. How will we bringing more relevance using responsible AI and we are creating a balance between the trust and customer trust and loyalty, conversions and optimization as what marketers say, and using Generative AI. So I would like to start with my first question and this is to do about relevance to customers and maybe I like to start with you, Kavita, is today I think we are reaching out to prospect customers through multiple channels. It's creating a lot of noise and to some level spamming as well. How do you maintain relevancy to customers, right? And how do you make sure that when we reach out to them, we reach out to them in the medium they would prefer and it's more relevant as a message. Sure, thank you for that question. So as you all would be aware, LNT Finance has a wide spectrum of products ranging from rural finance, micro loans, tractor finance as well as urban finance. So it's a really wide audience that we cater to. Also therefore, our mission is to leverage AI not just to sort of maximize reach, but also try to create, let's say customer delight, which is what is the right point in the consumer's journey that we need to sort of put our message out there. So we really look at the what, the who and the how of our campaigns. So to start with, in order to understand our consumers really well, we sort of assimilate the data on our consumers across touch points and really create those understanding of the individual preferences, try to find out which are the most effective ways to reach to the consumer, eliminate ineffective, let's say campaigns and create those tailor-made solutions for our audience. The other piece which is really important is to sort of understand the causality of consumer behavior. So we actually deep dive into understand why the consumer is behaving in a certain way and therefore target our communication in accordance to that. So a simple example would be that, let's say you have a credit score which all of us have to sort of understand the credit worthiness of our consumers. Now, let's say we know out of the search history of our consumers that all those who would have searched or found our information upon their credit score in the last three months have a higher propensity to consume alone. So therefore, using this causality of consumer behavior and then targeting creates that customer delight because they are in that journey of exploring let's say different solutions, loan solutions. And one such behavior that they exhibit is to really check on their credit score. And that's really our, let's say marker to know that if the consumer is searching their credit or is finding out about their credit score, there is a very high propensity for them to convert for alone. And that's the time that we then target our communication with the relevant messaging and all of this is obviously enabled because through AI we have assimilated the behavior of consumer across different touch points. Then of course is really the who which is using cohorts create those different personas of consumer and since I said we have a very large spectrum we have as high as 50 personas in the rural audience itself because that's the kind of behavioral pattern that each cohort really exhibits. And therefore as sharper we get in terms of defining these cohorts the better is obviously our campaign marketing. So that's really the parameters that we look at that can help to create customer delight. Yeah, no thanks. Thanks Kabita, I think this is interesting because you touched upon a very important point about reaching customers at scale relevance and you touched upon interacting with them in the dialects that they understand. I think India is a very diverse country, multiple dialects, literacy levels are very, very different. So I think you touched upon those areas. Hector, I think in the financial services space Tata Capital again you engage with customers with all spectrums, different demographic profiles. How do you see the reaching out to them with relevance? So I mean whatever Kabita spoke is quite frankly absolutely exactly applicable to Tata Capital also because we have all sorts of retail products, corporate products, we have an investment product. So I think what AI allows you to do is of course scale because there are lots of personas that you can work on like Kabita was saying and of course the Tata which you are reaching out I would not actually, let me take it back, not that the nimbleness with which you are replying is obviously comes from JNAI. And but we were discussing when we met with it that I think relevance is really, really underrated. Okay, when I say that I mean that even today Google is gonna crawl your content, for example, on relevancy. It's yes, you need to have enough keywords but in that I'm really oversimplifying here but what I mean to say is that if you implement JNAI and if you are able to generate 200 blocks instead of 100 doesn't mean you generate 200 blocks. The idea is to still generate 100 blocks with relevancy at the core of it. So I'm saying it does provide a lot of scale, does provide a lot of improvement on how soon we can solution things for our customers and definitely it does bring in a lot of relevancy not in all aspects, not in everything because it is largely prompt driven to a great extent but I think it provides us another world altogether compared to what we were in maybe a year ago. Well, thanks, thanks Hector. So moving on to I think the three panelists we have is from Life Insurance. I think one of the most toughest product to sell and you realize the value when you're no more. But interestingly, I would like to hear from you with Anjali that in the current scheme of things like when you reach out to customers and we heard, Kavita, we heard Hector from a relevancy standpoint making it more relevant across different channels. How do you simplify the message? How do you personalize the message and deliver at scale using Generative AI, right? And also make them understand, right, that okay and engage with them at scale because most of the interactions by the customers genuinely happens in my view at the claim processing level not at the sales level, right? So how do you bring that balance together? There are two areas that I'd like to touch upon. Of course the category is extremely complex as we all know, right? The products, the fact that we haven't even been able to scale our online marketing efforts is because of the fact that the customer needs even today a human interface to engage and converse to understand what is it that we have to offer and how is it that we can help them in their financial goals. Jenny I of course helps simplify that part because at scale it simply breaks down what is it that the customer can gain if his capacity is X versus another capacity, person's capacity is Y, right? Everybody's desires are different, life stages are different, input output of the household is different so it just breaks it down to that level. The other thing and I can speak for the entire category and my fellow panelists also here, we are a category that is plagued with a lot of misinformation. It's already complex and people take advantage of the fact because yes, it's not still a pull product, it's a push product. With this there is seamlessness of offerings, there is seamlessness of the communication that we have and the numbers specifically, which I think Jenny I is going to play a much larger role to bring ethics into the category. It is of course doing exactly what my fellow panelists Hector and Kavita also mentioned it's going to help us market years in prospecting and customer experience, but I think on a much bigger level it is going to help us perhaps in many ways build trust into the category. Yeah, yeah, well said Gita Anjali. I think this brings me one point of view. Most of the marketers out here in the room, I think we struggle to have the integration of generative AI with our legacy systems, right? How do we make sure that we bring the journeys together? Customer doesn't care whether you're reaching out to him at an acquisition level or at a service level. So Irem would you like to touch upon that? How do you resolving that entire conundrum about? Yeah, thanks Oded. So let's take a step back, right? Let us first understand what is AI going to do for the customers. We are talking from a perspective of marketers, which is obviously what we're talking about, but what does the customer expect? And the customer always expects that you are on time and personalized. So we speak about hyper-personalization, right? How is AI going to help me hyper-personalize my message and end to the lady who's my prospect? That is the key question. And I think everybody will here agree that AI is giving you unlimited scope of reaching your customers. First of all, it's three things, right? AI does three things. One, the amount of data with which we are grappling is humongous in all organizations. First and foremost, we must understand, and that's what we did in AG's federal life insurance. We used AI to first of all segregate and segment our customers, okay? I know my customers age, geography, where she's coming from, whether she's a mother or not. But I also know through various interactions how many times she has complained to me, how many times she has raised a ticket. What was the complaint? Was the complaint resolved or not? Was it resolved in time? What are the behaviors? I'm tracking her online behavior. AI helps me to segment her and compartmentalize customers like her into segments. Then what AI helps us to do is targeted communication. Marketing is all about communication. AI as a tool is helping us at AG's federal to target segment-wise communication. And so today, when I'm sending an online letter to my customers, at least 70, 17, 18 types of communications go to the same segments. So that's what AI is helping us to do. And the last bit, which is marketing and streamlining your campaigns. As marketers, we have always grappled with this notion that okay, I'll hire a media agency. I'll hire a creative agency to take care of my marketing needs. Yes, that is essential because they're the ones, they're the experts in their domains. But together with them, we have to keep this AI integration through various APIs to understanding of what is emerging in the market and make use of the AI possibilities to reach out to my customer with the right type of messaging. And that's what we're doing. So I think what we're doing, these three are the main functionalities and pros of AI, which everybody should utilize. I think brilliant. I think well said, Irem, it feels like marketing is less of a marketing job, but of an information officer now, right? Or a technology officer for that matter. I wish I, with that point, I want to touch upon, do you see any kind of challenges in terms of accepting AI internally in the organization or externally? How customers are reacting? How internally the teams are reacting? Do you see any kind of challenges from the AI standpoint? So coming last in the panel, especially when third in the life insurance category within the same panel, so I will not have too much to add over here. But the couple of challenges, first of all, I see. First of all, we need to make people understand what AI is. Exactly. We all keep on talking about AI, AI, AI. I don't think we all understand, even I will not understand, hand on hand, even I will not understand what exactly is AI, what does it do? What does it do, which was not being done earlier? So, subsequently finding acceptability, AI sounds very sexy to talk about, very good to talk about, but at the core what does AI do is very difficult to explain and that is the challenge that we find, that this is what AI does for you. The second part, the challenge that I see, I briefly touched upon very briefly in the last panel, but I'll probably talk a little more about that is, you know, one part in this, we on the screen saying one of the parts, the customer centricity, that's the key word over here. Unfortunately in our industry, and it's a little controversial statement, customer means and consumers, unfortunately, and I'm using the word unfortunately, because, and Gitanjali touched upon this, that bulk of the sale, 90, and I'm giving an industry figure, 92% of sales in this industry happens when there is an intermediary in between, that intermediary could be a life insurance agent, that could be a bank RM, in your case most of the time it'll be a bank RM, it could be a RM of a corporate agent, it could be a direct employee of yours, but there's always somebody between the insurer as an organization and the customer. And this interview, I will call this person a distributor for everybody's ease of understand. We all talk about customer centricity, customer centricity, think of a situation where the product that you're selling is complex, product that you're selling nobody wants to buy, nobody. In this room, nobody will get up in the morning and say, today I want to buy life insurance, nobody. So product is complex, you don't want to buy it, and third, which is the industry doing is, we have missold in past. As an industry, we have missold in past. And now thankfully, because of an extremely strong vigilant regulator, plus a lot of soul searching by the companies, misselling is very down to minimum, it still exists, but it's fairly low. So complexity, less trust and people don't want to buy. You combine these three and you have a product which is extremely difficult to sell. So therefore the person who's selling becomes so important to us. Unfortunately, customer centricity may, is person ka koi bhi, nobody thinks about this person. So the need for the AI, need for the new technologies, also that when you're developing this, when you're talking about this, think of this person as well. And I'm very specific to life insurance industry. So therefore, so what do we do about it? So we actually do, I would actually again, a little more controversial statement. We do more initiatives for JNAI towards the distributed side, rather than the consumer side. Because number one, it's more captive audience for me. So I can explain to them what I'm trying to do. Second is I can get them to experiment a lot because it is also an experiment. And third is we are learning. We do not know what this can do, what this cannot do, in which direction can it go. So because of the captive audience, I can do a lot of controlled experiments which we try to do with JNAI. And so therefore, a bulk of our focus is actually towards the distributed side. How can I create content for them which can at scale, at vernacular, you talked about vernacularization, Kavita talked about vernacularization. So how can I create content for them which they can reach out to customers? How I can help them to personalize their content for their customers. So that they can personalize it and then offer it to our customers. So a lot of work goes on that, or more towards that and less towards consumers. I think well said. And you brought a lot of reality. In fact, while I was listening to you, one of the use case comes into my mind is, I think there is a lot of fine print and the insurance policy or document goes through. I'm not sure any of you have actually read through a policy document. At least I haven't at any point in time. And even if I've tried to, I've never understood it. So using AI to really answer a query from that small LLM, I think one of the interesting research, not a research, but a use case, maybe an insurance industry is going through, actually create a micro LLM out there for policies that as a consumer, I can actually query and ask whatever I want to and I get an answer which is very fair and transparent, right? Sometimes you really don't understand what is the causation that would happen. Yes, Hiram, please. Yeah, so I'd like to, and very well said, I wish I, and in fact, he mentioned whatever the points he mentioned are true for most probably at some point or the other for all organizations. What I'd like to bring the discussion back to is the fact, what are we going to do as Abhishek also pointed out? What are we going to do with AI now at our disposal? Now I'll give you a use case from our company, Agis Federal. I don't know whether you guys know it, but our brand ambassador is Mr. Sachin Tendulkar, right? And he doesn't need any introduction. He's a Bharatsratan, God of cricket. What we did with AI and Sachin Tendulkar, eight month back, as you rightly pointed, nobody, once the premium of an insurance policy is paid, first of all, nobody understands the policy, then the biggest challenge which we all face is at the time of renewal in the next year, nobody either remembers to renew the policy or they are reluctant to renew the policy. What we did was we used generative AI, used Sachin Tendulkar, shot a small clip, where in Sachin Tendulkar is saying, so suppose if I'm the policy holder, Erum Kidwai gets a personalized WhatsApp message where Sachin is saying, hey, Erum Kidwai, thank you very much for being a customer of Agis Federal. And then he goes on to explain the benefits of renewal, timely renewal. Now just imagine, and when we conceptualized this concept, it was a novel idea. Sachin himself liked it a lot. Our CEO supported us full-fledgedly. Believe you me, we wanted this to be an initiative to increase our renewals, right? For more and more people to come and prepare renewals in time. It become, these videos went on to become viral videos because just imagine you're getting a message, Sachin is calling you by name and then telling you the benefits of paying renewal in time, right? And you're gonna share it with your entire group. So this is one use case which you can take up. We were the first ones to do, again, we were the first ones, at least in the BFSI segment in India, to use AI two and a half years back because we realized the potential that AI is one tool that will be able to flip your business to the positive side and take advantage of the technology rather than just wait for somebody else to take decisions and then copy it. So I think like the technology industry and like AI, it will now serve corporates to be proactive and innovative. You need to take that leap. And for that leap, as he rightly said, first of all, you need to understand what are the possibilities? Just by talking about them won't help, right? So I think that that's something which needs to be the norm of the day. Yeah, please. RM, you should share that video with me because a much younger cricketer who's our brand ambassador is hesitant to get on the generative AI bandwagon. So yeah. But in this whole mesh of like, as a marketer, we care about optimization. We care about conversions. But at the same time, we also care about, you know, not losing the customer, especially in the category that you belong, right? Also in the financial services space where you actually have the micro lending, right? And also you want to create maybe a term called brand loyalty, which perhaps don't exist so much now in these days in the age of AI and being so standardized thing. Where do you see the balance? Maybe any one of you, like Pitanj, do you want to add up and maybe Kapita then you can add up? Sure. So obviously, like I said previously, AI significantly helps us to drive brand loyalty. And one good use case for us is really, you know, improving, let's say the conversions amongst the first time loan applicants. Now, if you know a large part of India or Indians still don't have, let's say a credit score. And if I ask around this room as well, how many of you are aware of or have, you know, understanding of your credit score, you know, that is significantly solved by the use of AI. Because, you know, in the past, we would have a lot of first time loan applicants as high as 50% being rejected, you know, for their loan applicant application because of over emphasis on the credit history or lack of a credit score. Now, that's where AI has helped significantly because with the use of AI, we're able to solve this problem through predictive analysis, you know, through tracking the digital footprints of our consumers. And also to do what I say a good amount of segmentation. So, you know, observing the behavior of the consumers online, their online shopping, their social media profiles, or, you know, even the way they are sort of, you know, having their bills, utility bills paid out. Now, all of this helps you to assess, you know, the credit worthiness of an individual on a real time basis. In any typical scenario, let's say this person's loan would have been rejected, but all thanks to AI and the recent credit behavior, you know, you are able to therefore improve the probability of, you know, offering a loan. So I think that opens up a very large opportunity to really tap into the first time loan applicants who would have otherwise been rejected, but all thanks to AI, you know, there's a lot of credit worthiness that we can actually track on a real time basis. Yeah, thanks. So you mentioned about brand loyalty, Udit. I think it opens a different Pandora box because the spectrum is extremely wide, I mean, from deodorant to banking, right? The brand loyalty differs extremely in each of these categories. So I don't think it's always brand loyalty which is a challenge, especially in financial services because you stick to your RM and you have certain relationship and faith where you're giving your money to someone to invest and at the same time taking money and the ease of it, et cetera. So the parameters and the dynamics are extremely different. What I definitely think that organizations need to look within themselves, build a culture of upskilling themselves, creating a lot of understanding of it because in reality, our infosex still does not allow a lot of us to have open network and even have chat GPT and JNAI tools on our respective machines because there's a lot of unknown, as a lot of us also spoke about, that the areas that we do not know from an ethics standpoint, from a compliance standpoint, from a regulatory standpoint. So we ourselves need to, and I think it's a concern globally, otherwise we would have educated ourselves more but then again, we need to be more open. Even as organizations to understand the areas, it can impact whether it is analytics or it's customer service or it is campaigns. It has, it's a tool that is going to enable like calculators and fasten or go to market strategies in every sense. So education within the organization is extremely critical and it has to be top down driven. It has to be a conversation that is beyond just the marketing function and in permeate into every department of the organization. So I'm glad you touched upon it. I think that's one of the big internal challenges, right? When you actually embrace AI within the organizations. So having talk about AI and having to use AI and having to implement AI internally, I think there are three different things and you spoke about actually implementing AI, building that infrastructure and acceptability from various things. And I think this is bringing me to another question about responsible AI, right? I think we are all using, you all are into the regulated part, regulated businesses. So like Hector, we want to start with you. How do you see this whole space of responsible AI? Like for example, voice parts, I'm hearing about voice cloning for that matter, right? I'm hearing about many other instances about where AI could be not used very responsibly. DPDP bill here in India released, we are moving towards the GDPR era. So how do you see this whole balance between privacy and trust? So I think coming from what Abhishek spoke earlier, see we firstly need to choose as organizations what seems to be the use cases that we want to solve immediately through AI, right? And what seems like sort of a black hole probably needs to, you know, we need to wait it out. You know, till we have a certain confidence in that system. Secondly, see we were discussing this earlier also, I think use cases, there are a plethora of them. I don't think there is any scarcity of that. You can use it on, you know, for your online reputation management. You can use it on your web search tool. You can use it on your chat board. You can, I mean, it's endless. In fact, adding to what Abhishek was saying earlier, we've actually implemented it even for our credit managers. You know, it's underway, it's not implemented yet, but you know, which will reduce, this is on the corporate side, corporate lending, which will reduce the time taken for credit assessment dramatically. But again, right now it is within the realm of the information that is fed to JNAI, right? So therefore there is confidence that there is no question of any misuse. Having said that, we have to wait it out. I mean, I don't think everything has to be implemented immediately. Although we are excited, we are exploring all possibilities, voice to text, the other way around. I mean, but we have to hold on to our horses before we, you know, go crazy on it. Yeah, because this is a very interesting part, right? Coming from a regulated industry, it's not easy to implement everything, perhaps what could be easy for other industries, right? And with the AI regulation bill coming up in India, right? Abhishek, would you like to add something, like how will it really touch upon specifically customer interactions in the VFSI space? Yeah, so coming back to customer centricity, for a life insurance business, the ultimate measure of customer centricity that is very easy to understand by external people is a claim settlement ratio. You know, most of the time, many times, one of the deciding factors for people while choosing an insurance is the claim settlement ratio. So very broadly, if I see what is claim settlement ratio, number of claims that I have received versus number of claims that I have settled. Other claim settlement ratio is lower. Normal, the perception is that this probably company is rejecting a lot of claims. That is the perception. Now, how is AI helping in this? And let me give you an example. To answer that question, you have to first ask a question, why does a claim glare at reject? Correct. Does the insurance company really don't want to pay claims? No. All insurance companies, all insurance company wants to pay claims because that is the moment of truth. That is the reason why we are in business because when customer has decided to take a policy from us, customer has given the money, I have given only a piece of paper to the customer, which is a promise, which is that if you are not there, I am going to take care of the financial needs of your family. That's the promise. And therefore, all insurance companies wants to pay. Why does a claim get rejected? Claim gets rejected most of the time, only in a scenario when the customer, or without the knowledge of the customer, or the distributor, has filled in something in the application form, which is not true. And most of the time, these wrong information is non-disclosures. Non-disclosures regarding health, non-disclosures regarding other financial products that you have anything, but non-disclosure. So that means if I have an incoming policy which is a non-disclosure, when the time for claim comes, and if it is within three years, after three years you can't reject a claim, that's the law. Within three years, an insurance company can reject a claim. If a wrong policy has come in, and by the time the claim comes, and it is within three years, I as an insurance company will investigate. And if it was non-disclosure, I will have no other option but to reject it, indirectly impacting my claim settlement ratio, indirectly impacting my customer's entity. So how does I help in this? We have models running, that whenever a case comes in, I have my models running, these models have been perfected over a period of time. That depending on the parameters, I know that this cohort looks risky. Although I've agreed to underwrite this policy, but this cohort looks risky. So what we do, and in my previous panel I talked about that human connection, we then look at these cases case by case, and try to do a prior investigation so that a wrong policy does not enter. Wrong policy does not enter. What is my main motive? My main motive is only so that my claim settlement ratio earlier, later, could not get impacted, should not get impacted. As we say, garbage in, garbage out, that will happen, if you do not have these checks and balances in place. AI is a wonderful use case in this, wonderful use case. In fact, what we have done is, the successful model which has been for us, and we have been running it for close to two years now, it's so perfect, now it's close to, it works on a close to 90, 95% accuracy right now. If that model has thrown up a case, it is quite sure that this case might be something problem with that. What has that done? Obviously, there's a financial implication and it has done wonders for that. What it has done is, it has improved the acceptability of AI within the organization. So now we have teams coming up that I have this problem, can AI help me solve this problem? And a good part of a solution is always defining the problem. Classic days, when market research used to be really, really followed very rigorously, if you have defined the research objective very clearly, you will get a good output. If you have not defined, you will never get a good output. Similarly, is case with the AI. As long as you have been able to define your problem correctly, you will have a tool, you will have a model, you will have a working thesis which will come and give you a solution. Spend some time in defining the problem. Stay in the problem and you're going to get a fantastic output. Good, thanks Abhishek. Just a related question around Iram maybe or Gitanjali, anybody you can contribute. Abhishek, you talked about how AI is creating differentiation within the company, reaching out to the customers, creating more speed of execution, right? How is it, customers are accepting that, right? How is customer accepting to react with AI for that matter? We are all talking about the age of humanoids where humans and bots will come together. I hear about things like digital agents and everything. So how is customer acceptance coming in and there, and do you see a shift out there maybe and any light on that? Very valid question or that. So I'll split it into two parts, right? First is customer expectation, which I touched upon briefly in my opening statement. What does the customer expect from you today? My 13-year-old son, and I'll give you a very short story, eight months back, my 13-year-old son came to me, said, Papa, I've got a video to show you. I said, fine, show it to me. We are both football freaks, so I thought it would show me a football video. He showed it to me and there I was. I was, it was me, myself, speaking and saying that I should allow my son and my daughter more recreation time and not force them to study, right? Some generative AI app, he picked up and did this because he had my videos, he had my face, he had my voice. And it was a funny thing we all laughed, but you come to think of it, a 13-year-old person can do this to you free of cost. Now go back to his psyche. People, our customers, forget about 13-year-old, they are future customers. What will a 20, a 25-year-old, a 35-year-old, or even a 45-year-old intelligent customer expect from your brand? Will they not expect that you should have all touch points covered as a corporate, as a company? Forget about the industry also. That is the least expectation. When internet came into existence, after three years or four years, do you think that businesses anywhere across the world expected you to liars with them through letters or inland notes? No, they expected you to email them, right? That's the same expectation the customers will hold. I mean, needless to say, you don't need to think about that expectation. It has already, she has already started expecting you to use these new mechanisms and technological advancements to touch base with her, your customer. Now coming on to the second part. We always say hyper-personalization, right? AI and tools like ML, deep ML, deep learning, and algorithms-based models give you an opportunity to hyper-personalize because once you use AI mechanisms and you have API integrations with your in-built, in-house capabilities, what you'll do is you'll segment them, you'll give them target communication, you'll use their own data to reach out to them in a better manner and explain to them what they need to know from your services, from your offerings. So I think that is how we should, as corporates, look at it, and certainly as marketers because it's making my job easier and my team's job easier. I have more time to think and strategize rather than run around agencies and do the legwork. So I think that's the biggest advantage and we should exploit AI. Yeah, no, thanks. Gitaanji, would you like to add how Gen-Z is reacting to these AI-led interactions? I'll probably paraphrase what in fact Erem also mentioned. We're asking this question two years too late. We're in the world of Netflix and Amazon as consumers ourselves. We ourselves are extremely used to hyper-personalization. So I won't say consumers are reacting. I think they might be thinking in their heads that, hey, it's finally come to this category as well. So because clearly, I mean, we should have done this two years ago when it was still a big talk of the town when Cadbury launched the campaign which eventually got them cunts, which was two years ago. So it is not new, the subject is not new anymore. As organizations, we are late at certain categories. We are not only late, we are still very scared. And I think every passing day, we are becoming stale in this conversation and not being right in helping our businesses or the customer or the distributor in expediting the speed to go to market. COVID changed many things for us, right? We're talking of a generation that's, and many of us here are accomplice where we spend hours in doom-scrolling. And the algorithms there are also using the same tools as us, right? So we need content on the fly. I mean, there were times that organizations used to produce two to three films a year. That also is the largest number. Right now, that has come down to two to three films a week. And we have to account for time, we have to account for smartness using such platforms, we have to account for budgets. And yes, there will be agencies which will also have to help us and support us in this ecosystem because I will never have the kind of budget to produce this much content in a year, right? So we all will have to be extremely smart to play catch-up, to give the consumer what he needs and what he's been used to because the consumer doesn't care whether you're insurance or you're Amazon. I need the same experience from everybody that I'm engaging with. Yeah, no, thanks. I think we're just coming to the end. Before we conclude, I just want to ask, and this is kind of a one answer, like a rapid-fire kind of a thing, maybe, any generative AI or forget about generative AI, any AI tool that you're currently using to engage with customers or which is market-facing, which is not internal-facing, right? To improve the speed of execution, speed of response to customers or building more engagement with stickiness with the customers. Anything, any one tool that you must be using, Kavita? I think clearly in this age of video and image, all the newer technologies that are emerging there, so especially analyzing the video interactions where we do a lot of video KYC or video PDs, et cetera, with our consumers. And at that touch point, how is it that we can assimilate a lot more understanding about the ecosystem of our users? What habitat are they in? What would possibly be their lifestyle? And therefore, sharpen our, let's say, targeting to that user base just via that one interaction. So get a lot more out of that interaction to sharply target. Is one tool that we would use increasingly? We would use increasingly. So I think, like I said before, in possibly everything, I mean, all projects have not come to fruition, but everything, like from integrating JNAI on our ORM tool to doing that for our call center, to doing that, I gave you the example of credit managers very soon getting access to something which will turn their cycle of generating that report from probably one or two days to just 15 minutes, right? And doesn't mean his job is over, but the fact that he or she can rely on that to kind of look at that report within 15 minutes is going to give him or her so much more time to evaluate better rather than always there being a gun on their head to pass on cases quickly, right? So apart that email automation is obviously something for service that we've already implemented. Chatbot automation is already implemented. So I think lots of cases, but in images and videos, I personally feel that there is still some bit of lack of clarity in terms of who's responsibility it is and as far as copyright infringement is concerned. So we are using, but we are using with caution because we consulted legal as well as multiple tools and all of them had the same kind of commitment which is that you can, it's your right to use, but it does not specifically remove copyright infringement from the conversation. So which is why we are using, but with caution. Got it, got it. So in fact, I think it's a parallel of the category. Data security is extremely critical for us, both for ethical reasons and compliance reasons, right? We do not use anything that is available in the open internet. We have very strong NDAs with certain agencies where data is passed through APIs and not really punched into a chat box on open internet. So there are certain agencies. Where extremely robust IT integration has taken place to pass that data. Sure. So we've been using obviously AI chatbots and even email chatbots and stuff like that. We are very soon going to introduce something which is called, you know, media assist which is instead of going and giving your blood sample, we'll be just taking a video, a small video of two minutes and the AI video will throw up possibilities of certain diseases. And that will most probably help us to underwrite the policy in a much better manner and much faster manner. But having said that, we are very, very conscious about the fact, as Gitanjali pointed out, cyber security is at the epitome when it comes to decision making at our organization at Agus Federal. We do not want to compromise on anything which has got to do with the personal space of my customer. And that is going to be always the priority. And I mean, as a marketer, we did it two years back and very soon, I think next week, we're releasing a small, our mainline ad campaign with Sachin where you're gonna see Sachin Tendulkar who's a five-year-old, an eight-year-old, a 10-year-old, a 16-year-old when he made the debut at Lahore. And it's AI generated? No, I can't, that was the storyline but you'll get to see it soon. So, yes, using AI, but very, very responsibly is the key and mantra at Agus Federal. Nice, nice. Normal, as I said, as my fellow participants have said, data, privacy and security is very important, especially we come from a highly regulated industry, highly regulated. Within those boundaries, things like externally available or centrally available like EK, YCC, KYC, video verifications of customers, chatbots, whatever is available right now, we try to use all of them. Some we are successful, some we are not, but it's always a journey. You know, one of the interesting things happened. I asked the panel, give me one tool, right? And I was patiently listening. None of you actually mentioned one tool and this is very interesting. It's actually cannot be one tool. It's a combination of tools, right? And each one of you actually spoke about combination of things to solve a problem, right? And one of the interesting things that came out to me as a takeaway that each one of you is so concerned about the problem that you want to solve doesn't mean that you have to have a stack of AI to do it, right? And very focused on the responsible AI as well. My takeaway from this panel is like AI everywhere, you really can't escape, right? You're using AI for the ease of business and ease of customer doing business with you. And actually it feels like the future of customer and the future of customer conversation is like you engage with the customer with the context, right? So unless you have a context which is completely proper and AI derives in terms of segmentation, in terms of preferences, in terms of dialects, AI just helps you reach out in the right context. So thank you so much.